Sarah Baitzel

From InterSciWiki
Jump to: navigation, search
Mayatrip2008.JPG
Field photo.jpg

I am a second-year archaeology graduate student in the Department of Anthropology at UCSD. My research focuses on mortuary archaeology in the Andes, specifically the Middle Moquegua Valley. The research for my Masters Thesis compares demographic age profiles from various cemeteries associated with the Tiwanaku culture (AD 500-1000) to test for migration and general lifestyle patterns.

A289 Project - Finding a better model for Eff's study of Average Adult Female Contributions to Subsistence - pdf of the paper for EduMod In his article on Galton's Problem, Anton Eff revisits the issue of spatial and phylogenetic relationship of cultural traits and uses Moran's I to test for relatedness between cultures. The relatedness may to a certain degree account for the similarities between cultures and should caution anthropologists in their quest to make cross-cultural comparisons. For this project, I am using R to extend Eff's example of v826, but looking at a number of variables from the Standard Cross-Cultural Sample (a subset of Murdock's Ethnographic Atlas with many additional variables) and analyzing how, statistically, they account for and contribute to the degree of average adult female contributions to subsistence. To put it in an oversimplified manner, my goal is to increase the R-squared value by changing and substituting variables, maintaining the same number of variables but looking for possible functional and/or structural connections in the interpretation.

Sarah's project is one of the EduMod modules where the original study was used and modified, leaving a new example of a different model that other students can run or modify.

She studied a solution to Wikipedia:Galton's problem using E. Anthon Eff's

Comparative_research_tools#Anthon Eff's SAR Procedures Simultaneous AutoRegression]]
to find a better model to account for Average Adult Female Contributions to Subsistence, the dependent variable in the autocorrelation network effects regression model.